anyblok_pyramid_rest_api 0.7.0

Tools to build rest api's

Anyblok Pyramid Rest Api

The main idea is to provide facilities for building restful api that interacts with AnyBlok models through a CRUD like pattern.

As dependecy, it use Cornice for its validators and schema abilities and Marshmallow for schema definition, serialization and deserialization (we have since split this work to AnyBlok_Marshmallow).

• Free software: Mozilla Public License Version 2.0

• Documentation: https://anyblok-pyramid-rest-api.readthedocs.io

Features

• Incoming request validation through schema (validation before database query and ability to validate several parts of the request object)

• CRUD queries always with request.validated data

• Data deserialization for response through schema

• Easy CRUD resource declaration (map a model on an endpoint)

• Automatic schema generation based on models introspection

• Advanced collection filtering, ordering, paging (querystring validation through schema)

Todo

• Helpers for JsonSchema or Swagger

• Advanced introspection for api documentation generation

Request lifecyle

incoming request

-> validators -> deserializer (json to dict) -> schema load -> request.validated -> request.errors

request.validated

-> crud -> resulting records
-> deserializer (records to schema dump)
-> serializer (default pyramid / cornice dict to json serializer)
-> response

Author

Franck Bret

• [e-mail]

• https://github.com/franckbret

Contributors

Jean-Sébastien Suzanne

• [e-mail]

• https://github.com/jssuzanne

• https://github.com/AnyBlok

Credits

• Anyblok

• Pyramid

• Cornice

• Marshmallow

• AnyBlok_Marshmallow

License

This Source Code Form is subject to the terms of the Mozilla Public License, v. 2.0. If a copy of the MPL was not distributed with this file,You can obtain one at http://mozilla.org/MPL/2.0/.

Copyright (c) 2017, Franck Bret

Change Log

All notable changes to this project will be documented in this file. This project adheres to Semantic Versioning.

0.7.0 (2020-12-07)

• Added entries in the querystring:

  • primary-keys[key1:key2:…:keyn]=va1:va2:…:van,vb1:vb2:…:vbn

  • ~primary-keys[key1:key2:…:keyn]=va1:va2:…:van,vb1:vb2:…:vbn

  • composite-filter[key1:key2:…:keyn][op1:op2:…:opn]=va1:va2:…:van,vb1:vb2:…:vbn

  • ~composite-filter[key1:key2:…:keyn][op1:op2:…:opn]=va1:va2:…:van,vb1:vb2:…:vbn

0.6.0 (2020-05-15)

Added

• CrudResource can now allow calls to certain http verb without being logged.

  * allow_unauthenticated_user_to_access_to_all_verbs
  * allow_unauthenticated_user_to_access_to_collection_get
  * allow_unauthenticated_user_to_access_to_collection_post
  * allow_unauthenticated_user_to_access_to_collection_patch
  * allow_unauthenticated_user_to_access_to_collection_put
  * allow_unauthenticated_user_to_access_to_collection_delete
  * allow_unauthenticated_user_to_access_to_get
  * allow_unauthenticated_user_to_access_to_delete
  * allow_unauthenticated_user_to_access_to_patch
  * allow_unauthenticated_user_to_access_to_put

Refactored

• Updated the order_by[operator]=fieldname by order_by[fieldname]=operator because more than one order by can be send

  Warning

  the oldest order_by is deprecated and will be remove later

0.5.0 (2019-06-21)

Refactored

• Tag can now be grouped

  @Adapter.tags('green', 'blue')
  def tag_is_green_or_blue(self, querystring, query, tags):
      query = query.join(self.registry.Customer.tags, aliased=True)
      query = query.filter(self.registry.Tag.name.in_(tags))
      return query

• adapter_cls is renamed QueryStringAdapter. the compatibility is keeped and a DeprecationWarning is sent

0.4.0 (2019-01-03)

Added

• CrudResource.execute(name, kwarg) decorator to define Service on:

  • collection_path/execute/name: validator on the body

  • path/execute/name: validator on the path and the body

  @resource(collection_path='/foo', path='/foo/{bar}')
  class MyResource(CrudResource):
      @CrudResource.service('print', collection=True)
      def plop(self):  # /foo/execute/print
          # verb : POST (default), GET, PUT, PATCH, DELETE, ...
          # schema optional
          # query = self.get_querystring('rest action given to get_model')
          # body = self.body
          ...
  
      @CrudResource.service('print')
      def otherplop(self):  # /foo/{bar}/execute/print
          # verb : POST (default), GET, PUT, PATCH, DELETE, ...
          # schema optional for body, pathschema for path
          # body = self.body
          ...

• Collection views to update and delete collection defined by querystring:

  • collection_put

  • collection_patch

  • collection_delete

Refactored

• If a request.error is found during the execution of a view in the crud resource, then a registry.rollback will be done

• Collection post is waiting a list of dict, Now it is possible to create more than one entry

0.3.0 (2018-12-07)

Removed

• Compatibility with python 3.3

Fixed

• QueryString filters and tags are executed only one time by query

• in the context manager to save the error, the registry is now rollbacked

Refactored

• The error field is now the key of the errors description dict (All validation errors messages are now dict)

0.2.1 (2018-10-06)

Added

• MANIFEST.in file

Removed

• VERSION file

0.2.0 (2018-10-01)

Added

• context key in querystring. The goal is to add some informations to help custom filter and tag to build their query

Refactored

• Now the querystring desarializer use regex to get the informations

0.1.0 (2018-09-26)

Added

• CRUDResource class to define REST api

• service behaviours